Hepatic Encephalopathy

Hepatic encephalopathy (HE) is a brain dysfunction caused by liver insufficiency and/or portosystemic shunting (PSS); it manifests as a wide spectrum of neurological or psychiatric abnormalities ranging from subclinical alterations to coma.^[1] The adjective "hepatic" implies a causal connection to liver insufficiency and/or perihepatic vascular shunting — HE is, at its core, a diagnosis of exclusion built on that causal link, not a description of any altered mental status in a patient who happens to have liver disease (see Assessment & Diagnosis).^[1]

See this platform's Cirrhosis chapter for general cirrhosis staging (Child-Pugh, MELD), ascites management, and variceal-bleed prophylaxis — background this chapter assumes rather than re-explains. This chapter is the dedicated deep-dive on HE itself: staging, the precipitant-identification framework, and the lactulose/rifaximin protocol.

Classification

Every case of HE should be classified by all four of the following factors, repeated at relevant clinical intervals.^[1]

• Underlying disease: Type A (from acute liver failure — managed under acute-liver-failure protocols, not this chapter, given its distinct risk of cerebral edema and intracranial hypertension), Type B (predominantly from portosystemic shunting without cirrhosis), or Type C (from cirrhosis) — overwhelmingly the presentation encountered in ambulatory and community pharmacy practice, and this chapter's focus.^[1]
• Severity grade: the West Haven Criteria, below.
• Time course: episodic; recurrent (bouts within a 6-month interval or less); or persistent (behavioral alterations always present, interspersed with relapses of overt HE).^[1]
• Precipitating factors: nonprecipitated, or precipitated with the specific precipitant identified — see Risk Factors, where this framework is covered in depth.^[1]

Epidemiology

Overt HE (OHE) defines a decompensating event in cirrhosis, on par with ascites or variceal bleeding. Its prevalence at the time cirrhosis is diagnosed is 10-14% overall, 16-21% in decompensated cirrhosis, and 10-50% after a transjugular intrahepatic portosystemic shunt (TIPS) procedure; cumulatively, OHE occurs at some point in 30-40% of people with cirrhosis, usually repeatedly.^[1] Minimal or covert HE — abnormal only on specialized testing, with no clinical sign — occurs in 20-80% of people with cirrhosis, reflecting real variation in which testing tool is used.^[1] After a first OHE episode, the cumulative risk of recurrence is 40% at 1 year; after a recurrent episode, it is 40% within 6 months despite lactulose therapy — the recurrence-prevention protocol in Choose Initial Treatment exists precisely because of this high baseline recurrence risk.^[1] HE accounted for approximately 110,000 hospitalizations per year in the United States (2005-2009).^[1]

The clinical spectrum

In its mildest expression, HE alters only specialized psychometric tests of attention, working memory, psychomotor speed, and visuospatial ability, with no clinical sign at all.^[1] As it progresses, family members often notice personality change — apathy, irritability, disinhibition — and a reversed or disrupted sleep-wake cycle before the patient or a clinician does; progressive disorientation, inappropriate behavior, agitation or somnolence, stupor, and finally coma follow.^[1] Asterixis ("flapping tremor") is a negative myoclonus — a brief loss of postural tone, not a true tremor — classically elicited by hyperextending the wrists with the fingers spread; it is not specific to HE (it also occurs in uremia) and is often present in the early-to-middle stages preceding stupor or coma.^[1] Extrapyramidal findings (hypomimia, rigidity, bradykinesia, a parkinsonian-like tremor) are common; deep tendon reflexes and motor/mental signs do not necessarily progress together in a given patient, which genuinely complicates bedside staging.^[1]

West Haven Criteria — the central staging framework

Overt HE (OHE) is diagnosed by clinical criteria and graded using the West Haven Criteria (WHC), the gold standard, alongside the ISHEN covert/overt correlation. The recognition of disorientation and asterixis specifically has good inter-rater reliability and are used as OHE's marker symptoms; Grade I findings are the most easily overlooked on a routine exam.^[1]

All conditions above are required to be related to liver insufficiency and/or portosystemic shunting.^[1] In significantly altered consciousness, the Glasgow Coma Scale supplies an additional operative description alongside the WHC.^[1]

Reversibility, and the real long-term stakes

It remains a well-founded operational basis for treatment that HE is reversible — but research in liver-transplant recipients and in patients after repeated OHE bouts has genuinely cast doubt on complete reversibility: some mental deficits can persist beyond what is explained by other transplant-related causes, and episodes of OHE have been associated with cumulative, persistent deficits in working memory and learning.^[1] Even minimal/covert HE meaningfully reduces quality of life and is itself a risk factor for a future OHE bout, and can affect driving and occupational safety despite a normal-appearing routine exam — see Patient Education in Non-Pharmacologic Treatment and the diagnostic-reevaluation point in Common Errors to Avoid.^[1]

Risk factors for a first OHE episode

Within 5 years of a cirrhosis diagnosis, the risk of a first OHE bout is 5-25%, depending on the presence of other cirrhosis complications (prior minimal/covert HE, infection, variceal bleeding, ascites) and, probably, diabetes and hepatitis C.^[1] Even people with cirrhosis and only mild cognitive dysfunction or mild EEG slowing develop, on average, approximately one OHE bout per 3 years of survival.^[1] See the Cirrhosis chapter for the underlying causes and general risk factors for cirrhosis itself.

^[1] Each of these precipitants is genuinely pharmacist-actionable:

• Infection, including spontaneous bacterial peritonitis (SBP) — the single most frequent precipitant of episodic OHE; keep a low threshold for a diagnostic work-up (blood/urine culture, chest imaging, diagnostic paracentesis) rather than assuming the HE itself explains the whole picture. See the Cirrhosis chapter for SBP diagnosis and prophylaxis.
• GI bleeding — variceal or non-variceal; increases the nitrogen/protein load delivered to the gut and colon. Treat as a medical emergency in its own right (see Red Flags & Refer); see the Cirrhosis chapter for variceal-bleed management and prophylaxis.
• Constipation — increased colonic transit time favors bacterial ammonia production and its colonic absorption; this is the direct mechanistic rationale for lactulose itself (see Choose Initial Treatment).
• Dehydration and diuretic overuse — reduces renal ammonia clearance and hemoconcentrates; review the diuretic regimen alongside the Dehydration in Adults chapter's sick-day medication guidance.
• Electrolyte disorder, especially hypokalemia — hypokalemia (often itself diuretic-related) increases renal ammonia generation and favors ammonia's un-ionized, brain-permeable form; correcting potassium is a genuine, specific treatment action, not just routine lab monitoring.
• Medication nonadherence with HE therapy — a missed lactulose or rifaximin dose is one of the most directly preventable precipitants; reinforce adherence explicitly at every relevant encounter (see Non-Pharmacologic Treatment).
• Sedative, opioid, or benzodiazepine use — direct CNS depressant effect that can precipitate or mask worsening HE, and independently raises aspiration risk in a patient with an already-impaired level of consciousness; reassess necessity of any new such agent (see Common Errors to Avoid).
• TIPS procedure — creates a portosystemic shunt that bypasses hepatic ammonia clearance; notably, routine prophylactic lactulose or rifaximin before or after TIPS is not recommended (see Special Populations and Choose Initial Treatment).^[1]

Diagnosis is clinical, and by exclusion

OHE is diagnosed by clinical examination and clinical judgment, not by a laboratory threshold; specific quantitative psychometric/neurophysiological testing (the PHES battery, Critical Flicker Frequency, and similar tools) is used mainly to detect minimal/covert HE and in study settings, not to diagnose an overt episode.^[1] Because patients with advanced cirrhosis are genuinely susceptible to mental-status abnormalities from other causes (medications, alcohol, other drug use, hyponatremia, psychiatric disease), HE remains a diagnosis of exclusion — recognizing a plausible precipitant (infection, bleeding, constipation) supports the diagnosis, and exclusion of other etiologies by laboratory and radiological assessment is warranted as clinically indicated.^[1]

Differential diagnosis of altered mental status in a cirrhotic patient

Hyponatremia and sepsis are each capable of producing encephalopathy in their own right and of precipitating HE through the mechanisms above — the two are not mutually exclusive, and in end-stage liver disease uremic encephalopathy and HE can overlap.^[1]

The role — and real limits — of blood ammonia

Increased blood ammonia alone does not add diagnostic, staging, or prognostic value for HE in patients with known chronic liver disease; conversely, a normal ammonia value in a patient with presumed OHE calls for diagnostic reevaluation of the HE diagnosis itself.^[1] This is echoed directly in Canadian practice: the Canadian Association for the Study of the Liver's own Choosing Wisely Canada recommendation states that clinicians should not order serum ammonia to diagnose or manage HE, since an elevated level adds no diagnostic, staging, or prognostic value in a patient already known to have chronic liver disease.^[2] The one narrow exception: in a patient already established on ammonia-lowering therapy, a repeated ammonia measurement can occasionally help gauge whether that specific drug is having a biochemical effect — a different question from using ammonia to diagnose or grade an episode, and not a routine practice.^[1] See Common Errors to Avoid.

Brain imaging and basic labs

CT, MRI, or other structural brain imaging do not contribute diagnostic or grading information for HE itself, but a scan is usually part of the work-up for first-time HE or whenever another intracranial pathology is suspected, given the substantially increased intracranial hemorrhage risk in cirrhosis noted above.^[1] Basic work-up to support the precipitant search in Risk Factors: CBC, electrolytes (with particular attention to potassium), renal function, glucose, INR/liver panel to reassess severity, and blood/urine cultures or chest imaging when infection is suspected.

• Reduce the burden of ammonia and other gut-derived neurotoxins reaching the brain.
• Resolve the acute overt episode and restore the patient's baseline level of consciousness and function.
• Identify and treat the precipitating factor — a co-equal priority to HE-specific drug therapy, not a secondary step.
• Prevent recurrence through appropriate secondary prophylaxis once an episode has occurred.
• Maintain adequate nutrition and lean body mass — never restrict protein to manage HE (see the Cirrhosis chapter and Non-Pharmacologic Treatment).
• Protect patient safety: airway/aspiration precautions during an episode, and fall/driving/occupational-safety counselling once resolved.
• Refer for liver transplant evaluation at the correct trigger point — recurrent intractable OHE with liver failure is itself an indication (see the Cirrhosis chapter).

Adjunctive and alternative options

• Polyethylene glycol 3350 (PEG) electrolyte solution — a single-center randomized trial found a 4-L PEG regimen produced faster and more complete resolution of an acute OHE episode than standard lactulose (91% vs. 52% improved at 24 hours), a genuinely notable result, though current guidance still describes PEG as needing further validation before it displaces lactulose as first-line; a reasonable alternative particularly when lactulose is not tolerated or not producing the expected bowel response.^[4][1]
• Oral branched-chain amino acids (BCAA) — can be used as an alternative or additional agent in patients not responding to conventional therapy, or to help meet nitrogen/protein targets in patients intolerant of standard dietary protein; intravenous BCAA has no demonstrated benefit for an acute HE episode.^[1]
• Intravenous L-ornithine L-aspartate (LOLA) — an alternative or additional agent for patients nonresponsive to conventional therapy; oral LOLA supplementation is not effective.^[1]
• Neomycin or metronidazole — legacy antibiotic alternatives for overt HE, still used by some clinicians, but long-term ototoxicity, nephrotoxicity (neomycin), and neurotoxicity (metronidazole) make both unattractive for continuous, long-term use now that rifaximin is available.^[1]

Discontinuing secondary prophylaxis

There is a near-uniform practice of continuing HE therapy indefinitely after a bout resolves, but this is not always necessary: if the recurrent precipitating factor has been durably controlled (e.g., an infection resolved, a variceal bleeding source treated) or if liver function/nutritional status has meaningfully improved, prophylactic lactulose or rifaximin may reasonably be discontinued on a case-by-case basis.^[1] Testing positive for minimal/covert HE before stopping therapy identifies patients at higher risk of recurrence and argues for continuing (see Monitoring & Follow-Up).^[1]

TIPS is the one setting where routine prophylaxis is not recommended

Do not start routine prophylactic lactulose or rifaximin specifically to prevent post-TIPS HE — a randomized trial found neither performed better than placebo for this indication.^[1] Manage post-TIPS HE reactively if it occurs; see Special Populations and the Cirrhosis chapter for shunt-diameter reduction in intractable cases.

Nutrition — never restrict protein

See the Cirrhosis chapter for full nutrition guidance; the HE-specific point bears repeating because it is a genuinely persistent and harmful misconception: chronic protein restriction does not reliably improve HE and instead worsens the sarcopenia and malnutrition that independently predict poor cirrhosis outcomes.^[1] Some brief protein restriction may be unavoidable in the first few days of a severe acute OHE episode, but should not be prolonged; if protein is genuinely not tolerated, substituting milk-based or vegetable protein, or supplementing with BCAA, is preferable to reducing total protein intake.^[1] Daily energy intake of 35-40 kcal/kg ideal body weight and protein intake of 1.2-1.5 g/kg/day are the targets, delivered as small meals evenly spread through the day plus a late-night snack to offset the accelerated overnight catabolic state of cirrhosis.^[1]

Supportive care during an episode

• Assess swallowing and aspiration risk before resuming oral intake in a patient recovering from Grade III-IV HE; a nasogastric tube can be used to give oral HE therapies to a patient unable to swallow safely or at risk of aspiration.^[1]
• Position and monitor patients with reduced consciousness for aspiration risk; ensure a documented plan for airway management escalation if the grade worsens (see Red Flags & Refer).
• Confirm the medication list has been reviewed for any newly started sedative, opioid, or benzodiazepine, and that it has been discontinued or reduced where clinically possible.

Patient and caregiver education

• Teach the early-warning signs, not just the dramatic ones — subtle personality change, altered sleep-wake pattern, or handwriting/coordination changes are often noticed by family before the patient or a clinician; encourage contacting the care team early rather than waiting for overt confusion.
• Explain the stooling target concretely — "2 to 3 soft bowel movements a day" is a self-monitorable instruction patients and caregivers can act on between visits; explain both under- and over-treatment (too few stools means the dose is too low; watery diarrhea, cramping, or new confusion between visits means the dose is too high).
• Reinforce adherence to lactulose and rifaximin explicitly — nonadherence is one of the most directly preventable precipitants of recurrence.
• Counsel patients and families to check with the pharmacist or prescriber before starting any new sedative, opioid, or over-the-counter sleep aid.
• Reinforce alcohol cessation (see the Cirrhosis chapter).
• Driving and occupational safety: even minimal/covert HE can affect driving and work performance despite a normal-appearing routine exam; a diagnosis of minimal/covert or overt HE is not automatically disqualifying, but pharmacists and prescribers should counsel patients on the possible risk and, in difficult cases, involve the authorities with expertise in fitness-to-drive assessment rather than making that determination alone.^[1]
• Basic fall-prevention and home-safety measures for any patient with residual cognitive impairment between episodes.
• Confirm the patient/family knows the Red Flags & Refer criteria for seeking urgent care.

Patients unable to take or tolerate oral lactulose

For patients with severe grade HE, ileus, or a genuine aspiration risk who cannot safely take oral or nasogastric lactulose, a lactulose retention enema is an accepted alternative route. A randomized study of rectal lactulose given alongside standard oral therapy used 300 mL of lactulose diluted in 700 mL of water, retained per protocol and repeated approximately every 4 hours, targeting the same 2-4 soft-stool response.^[5] Confirm local institutional protocol, since exact dilution/retention-time practice varies.

TIPS

Do not start routine prophylactic lactulose or rifaximin specifically to prevent post-TIPS HE — trial evidence does not support it over placebo for that purpose.^[1] If HE develops or persists after TIPS, manage it with the same precipitant-screening and lactulose/rifaximin approach used elsewhere in this chapter; for HE that remains intractable, shunt-diameter reduction is a specialist-directed option covered in the Cirrhosis chapter.

Renal impairment

Lactulose is not renally cleared and requires no renal dose adjustment; rifaximin's minimal systemic absorption likewise means it generally needs no renal adjustment. Watch for electrolyte and volume shifts from large-volume oral regimens (lactulose or PEG) in a patient with reduced renal reserve.

Older adults / frailty

Lactulose-related dehydration and falls carry disproportionate consequences in a frail older adult; titrate more conservatively and reassess more frequently. See the Dehydration in Adults chapter for orthostatic-vital-sign assessment technique if dehydration is suspected during lactulose titration.

Liver transplant candidates and recipients

Recurrent, intractable OHE together with liver failure is itself an indication for liver transplant evaluation (see the Cirrhosis chapter). HE should improve after transplant, but not all manifestations are guaranteed to fully reverse; large portosystemic shunts should be identified and embolization considered before or during transplantation, since they can cause persistent post-transplant HE. Distinguishing HE from a neurodegenerative disease (Alzheimer disease, cerebrovascular disease) matters more, not less, in a transplant candidate, since HE should improve after transplant while a neurodegenerative disorder will not.^[1]

Pregnancy and breastfeeding

Lactulose is a nonabsorbed synthetic disaccharide and is considered compatible with pregnancy and breastfeeding. Rifaximin's minimal systemic absorption suggests low fetal/infant exposure, but pregnancy-specific outcome data are limited — use per specialist guidance. Decompensated cirrhosis with HE is uncommon in pregnancy; manage jointly with obstetric and hepatology care (see the Cirrhosis chapter's pregnancy section).

• Track bowel-movement frequency at every relevant contact — this is the concrete, actionable lactulose-titration marker, not a background detail; confirm the patient is tracking toward, and not past, the 2-3 soft stools/day target.
• Re-grade using the West Haven Criteria at each relevant contact and with any suspected breakthrough episode, to track trajectory and to document the case fully (type, grade, time course, precipitated or not — see Clinical Snapshot).
• With any breakthrough or recurrence, reassess for an unaddressed precipitant from the Risk Factors list before empirically escalating the lactulose dose.
• Do not use serum ammonia to guide or monitor routine HE control — the one narrow exception is using repeated ammonia measurements to gauge an individual patient's biochemical response to an ammonia-lowering drug, not as a routine follow-up test.^[1]
• Periodically reassess whether ongoing secondary prophylaxis is still needed — it may reasonably be discontinued once the precipitant is durably controlled or liver/nutritional status has meaningfully improved (see Choose Initial Treatment).^[1]
• Consider minimal/covert HE testing before stopping HE drug therapy where available — a positive result predicts a higher risk of recurrence and argues for continuing rather than stopping.^[1]
• Monitor renal function and electrolytes, especially potassium, given hypokalemia's dual role as both a precipitant and a consequence of diuretic therapy.
• See the Cirrhosis chapter for the broader routine surveillance schedule (labs, HCC/variceal surveillance) that continues alongside HE-specific follow-up.

1. Lactulose not achieving the 2-3 stools/day target, or mental status not improving — first confirm dosing and adherence, then actively search for an unrecognized or inadequately treated precipitant (Risk Factors) rather than escalating the lactulose dose further; a lack of effect at an adequate dose is not remedied by a much larger dose.^[1]
2. Recurrent OHE (a second or later episode) despite adequate lactulose — add rifaximin 550 mg PO twice daily to lactulose (see Choose Initial Treatment).^[3]
3. Recurrent or breakthrough OHE despite lactulose plus rifaximin — consider oral BCAA or IV LOLA as an additional agent; reassess for a large spontaneous or TIPS-related portosystemic shunt that may warrant specialist-directed embolization or shunt-diameter reduction (see the Cirrhosis chapter), or for an unaddressed precipitant.^[1]
4. Grade worsening despite treatment (e.g., Grade II progressing to Grade III/IV) — treat as a red flag: reassess airway safety, and search harder for an alternative or additional cause of altered mental status (see Assessment & Diagnosis) rather than assuming simple HE progression.
5. Normal ammonia in a patient with presumed OHE that is not responding as expected — question the HE diagnosis itself and pursue the differential diagnosis work-up.^[1]
6. Recurrent, intractable OHE together with liver failure — this is itself an indication for liver transplant referral (see the Cirrhosis chapter) rather than continuing to escalate outpatient HE therapy indefinitely.^[1]

• Ordering or relying on a serum ammonia level to diagnose, stage, or guide day-to-day management of HE — the Canadian Association for the Study of the Liver's own Choosing Wisely Canada recommendation specifically advises against this; a normal level should prompt diagnostic reevaluation, not reassurance.^[2][1]
• Assuming an OHE episode occurred "spontaneously" without systematically screening for a precipitant — one can be identified in nearly all episodic bouts, and correcting it alone resolves about 90% of them.^[1]
• Escalating the lactulose dose indefinitely in a patient who isn't responding, instead of searching for an unrecognized precipitant or reassessing the diagnosis.^[1]
• Over-titrating lactulose past the 2-3 soft-stool/day target — this risks dehydration, hypernatremia, and aspiration, and can itself precipitate or worsen HE.^[1]
• Using rifaximin as a substitute for lactulose rather than an add-on to it — no solid evidence supports rifaximin monotherapy.^[1]
• Restricting dietary protein to manage HE — outdated, does not reliably help, and worsens sarcopenia and malnutrition (see the Cirrhosis chapter); brief restriction in the first days of a severe acute episode is the only exception and should not be prolonged.^[1]
• Starting routine prophylactic lactulose or rifaximin specifically to prevent post-TIPS HE — not supported by trial evidence.^[1]
• Continuing a newly started sedative, opioid, or benzodiazepine without reassessing necessity in a patient with any grade of HE.
• Dismissing minimal/covert HE as clinically unimportant because the patient "looks fine" on a routine exam — it carries real driving/occupational-safety and recurrence-risk implications.^[1]
• Defaulting to HE as the explanation for altered mental status in a cirrhotic patient without working through the differential diagnosis (hypoglycemia, alcohol withdrawal/Wernicke, sedative effect, intracranial hemorrhage, nonconvulsive seizure, and others).^[1]

How to use this table: Doses are adult doses and require patient-specific confirmation (renal/hepatic function, weight, allergy history). Pediatric dosing is weight-based per the current product monograph unless stated. Drug pricing is intentionally omitted. Select any drug name to open its complete Pharmasist Drug Database page.

Nonabsorbable Disaccharide — First-Line

Titrated to a stooling target, not a fixed dose. See Choose Initial Treatment for the full induction/maintenance protocol.

Antibiotic Add-On for Secondary Prophylaxis

Added to, never substituted for, lactulose.

Adjunctive and Alternative Agents

Reserved for patients not responding to lactulose ± rifaximin, or when lactulose is not tolerated.